Probably the top two mistakes I see people both here on AE and out in the industry make are not putting enough time into what they make, and not putting enough research into what they make.
In a nutshell I tell people to make your thing, then leave it for a day and come back to it later, does wonders for getting different iterations. The more iteration a design has gone through the better it ends up being. Related to this I must stress the importance of sitting down with a pencil and paper and getting your ideas out on paper first. Doesn't matter if it's a sketch of a logo, or a livery, or just writing out what ideas you might have for the work. It's a lot quicker than sitting in front of a computer screen and trying to draw it out in one go.
Precedent is also very important for making something believable, especially you folks taking up historical liveries but even for people doing contemporary liveries. It's also good if you're in a bit of a jam in terms of ideas. But yeah there's clever and innovative originality, then there's s***ty originality. Research guards against s***ty originality.
Personally I think nailing down these two skills is where people should start before they do anything else. Gets you into the right mindset to actually start designing good work.